Yokogawa In-Situ Gas Analyzer TDLS200 - Manual (Page 7)

7
IM11Y01B01-12E-A
1st Edition
Window Purge Gas Flow Adjustment – While there is no specific formula for this purge gas flow
rate (due to the many variables/complexity), we include the following information that might help
optimize the window purge gas flow rates actually at site. As mentioned in the User’s Guide, the
window purge flow rate can vary by application from as little as 5 lts./min up to as much as 50
lts./min for typical process applications.
There are a couple of methods that can be used to help establish a suitable flow rate.
For relatively short optical path lengths (such as process O2 measurements) if there is too much
flow the purge gas will protrude/mix into the process gas slightly and therefore possibly cause a
slightly shorter actual optical path length (readings lower than expected). Conversely, if there is too
little window purge flow the process gas may protrude/mix with the purge gas (worst case
contaminate the windows) in the connection nozzle and possibly cause a slightly longer actual
optical path (readings higher than expected).
One method that can be used (if the process gas is not excessively dirty or wet) would be to
measure the process gas without purge for a short period of time. Before starting this test, the
process measurement (O2 concentration) should be relatively stable and not expected to change
within a few minutes. First, establish the distance between the launch unit window and the detect
unit window and enter this value as the Process Path Length (under Advanced Menu, Configure).
Then stop the process window purge gas flow to both the launch and detect units while observing
the transmission – note, if there is a sudden large decrease in transmission then probably
contamination/fouling has occurred on the windows! Wait for the process reading to stabilize (i.e.
the process gas has filled the nozzle and alignment bellows sections up to each window), take note
of the measurement value and resume some window purge flow. Change the Process Path Length
back to the normal value and now resume/adjust the launch and detect unit window purges until
the process reading matches the value noted when there was no purge flow. If the transmission
does not return to its original value then the process windows will have to be cleaned to ensure
optimal operation. It is best to repeat this procedure two or three times to ensure repeatable results
that establish confidence in the readings. This can be repeated at any time at a later date to verify
the process readings.
A similar method can also be used by comparing ratios of the path lengths and measurements with
and without window purge flow. Using the analyzers integral Trend function, observe the
measurement values when both launch and detect window purge gas flows are simultaneously
stopped (for approx. 30 seconds or until the reading stabilizes) and then resume window purge
flow back to the initial flow rates. As with the other method, observe the transmission – note, if
there is a sudden large decrease in transmission then probably contamination/fouling has occurred
on the windows! Once the process measurement has re-stabilized with window purge gas flow,
repeat the simultaneous stopping of window purge gas flow. Repeat this several times until the
trend screen clearly shows repeatable measurement results with and without window purge. If the
transmission does not return to its original value then the process windows will have to be cleaned
to ensure optimal operation. The ratio of these two measurements can now be established based
on the repeatable results obtained. Establish the distance between the launch unit window and the
detect unit window and calculate the ratio with respect to the normal Process Path Length. The
ratio of ‘measurement without purge gas to measurement with purge gas’ should be equal to ‘path
length window-window to process path length’ – thus indicating that the connection nozzles are
purged correctly.